A model for process representation and synthesis

The problem of representing groups of loosely connected processes is investigated, and a model for process representation useful for synthesizing complex patterns of process behavior is developed. There are three parts, the first part isolates the concepts which form the basis for the process representation model by focusing on questions such as: What is a process; What is an event; Should one process be able to restrict the capabilities of another? The second part develops a model for process representation which captures the concepts and intuitions developed in the first part. The model presented is able to describe both the internal structure of individual processes and the interface structure between interacting processes. Much of the model's descriptive power derives from its use of the notion of process state as a vehicle for relating the internal and external aspects of process behavior. The third part demonstrates by example that the model for process representation is a useful one for synthesizing process behavior patterns. In it the model is used to define a variety of interesting process behavior patterns. The dissertation closes by suggesting how the model could be used as a semantic base for a very potent language extension facility.